debug
Pass
Audited by Gen Agent Trust Hub on May 18, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill is designed to run arbitrary shell commands inside a remote Docker container using the
tools/debugger/client.shscript. This functionality is the core purpose of the skill as a debugger. - [PROMPT_INJECTION]: The skill creates a surface for indirect prompt injection because it accepts natural language instructions to execute commands without explicit sanitization or boundary markers mentioned in the markdown body.
- Ingestion points: User instructions provided to the agent in natural language (e.g., "run in Docker ").
- Boundary markers: None specified in the SKILL.md instructions to separate instructions from data.
- Capability inventory: The skill enables full shell command execution via
bash tools/debugger/client.sh run "<command>". - Sanitization: No input validation or sanitization routines are described in the provided skill content.
Audit Metadata